Length-based methods have been widely applied to estimate biological parameters and understand the dynamics of marine resource populations within data-limited stocks. However, to date few studies have tested the sensitivity of parameters in length-based methods examining stocks with different traits and fishery contexts. In the Bay of Biscay and the Iberian Coast ecoregion, SE Europe (International Council for the Exploration of the Sea - ICES, Divisions 8 and 9a), many commercial resources are considered data-limited, and either little is known about their fisheries stock statuses or gaps remain in currently available assessments. Therefore, this study focuses on two of the most used length-based methods, which ICES considers to be the most appropriate to evaluate data-limited stocks, namely length-based indicators (LBI) and the length-based spawning potential ratio (LBSPR). Both methods have been applied to assess the stock statuses of various relevant species in the study area, such as: the small-spotted catshark (Scyliorhinus canicula), European anchovy (Engraulis encrasicolus), blackspot seabream (Pagellus bogaraveo), pouting (Trisopterus luscus), pollack (Pollachius pollachius), and Norway lobster (Nephrops norvegicus). For each stock, model results were compared with available knowledge on their current status. Furthermore, this paper discusses whether unexpected results are related to violations of the main model assumptions (constant total mortality and recruitment, and logistic selectivity) or to a lack of representativeness of stock length composition. A sensitivity analysis was conducted on the two most important input parameters: L∞ (von Bertalanffy asymptotic average maximum body size) and M/k (ratio of natural mortality to von Bertalanffy growth rate). This analysis concluded that the variation/misspecification of both parameters (M/k and L∞) had a considerable impact on the results given by both methods, and that this effect is more significant in the case of the L∞ parameter, thus highlighting the importance of its accuracy in a given assessment. Furthermore, the sensitivity analysis indicated that, among the LBI and LBSPR indicators, the least robust indicator was the LBI Pmega indicator. The remaining LBI method indicators can be considered more robust than the LBSPR indicators when there is uncertainty in the life history input parameters (M/k and L∞). Among LBSPR indicators, SPR is more affected than F/M (relative fishing mortality) by the variation/misspecification of parameter L∞, whereas under the variation of M/k both indicators perform similarly. However, it is important to consider that LBI indicators are very rough measures of stock statuses, whereas LBSPR measures describe stock statuses more explicitly. Thus, both measures can be used together to obtain more precision in estimating stock status. Nevertheless, when critical parameters are uncertain (e.g., L∞) and the results of both methods contradict one another, LBI method indicators, with the exception of Pmega, are more trustworthy than LBSPR indicators.
Read full abstract